Online clustering via finite mixtures of Dirichlet and minimum message length

被引：36

作者：

Bouguila, N ^{[1
]}

Ziou, D ^{[1
]}

机构：

[1] Univ Sherbrooke, Fac Sci, Dept Informat, Sherbrooke, PQ J1K 2R1, Canada

来源：

ENGINEERING APPLICATIONS OF ARTIFICIAL INTELLIGENCE | 2006年 / 19卷 / 04期

基金：

加拿大自然科学与工程研究理事会;

关键词：

online clustering; dirichlet; mixture modeling;

D O I：

10.1016/j.engappai.2006.01.012

中图分类号：

TP [自动化技术、计算机技术];

学科分类号：

0812 ;

摘要：

This paper presents an online algorithm for mixture model-based clustering. Mixture modeling is the problem of identifying and modeling components in a given set of data. The online algorithm is based on unsupervised learning of finite Dirichlet mixtures and a stochastic approach for estimates updating. For the selection of the number of clusters, we use the minimum message length (MML) approach. The proposed method is validated by synthetic data and by an application concerning the dynamic summarization of image databases. (C) 2006 Elsevier Ltd. All rights reserved.

引用

页码：371 / 379

页数：9

共 27 条

[1] NEW LOOK AT STATISTICAL-MODEL IDENTIFICATION [J].

AKAIKE, H .

IEEE TRANSACTIONS ON AUTOMATIC CONTROL, 1974, AC19 (06) :716-723

[2]

[Anonymous], 2000, FINITE MIXTURE MODEL

[3]

[Anonymous], ACM MULTIMEDIA

[4] Finding overlapping components with MML [J].

Baxter, RA ;

Oliver, JJ .

STATISTICS AND COMPUTING, 2000, 10 (01) :5-16

[5] Inference in model-based cluster analysis [J].

Bensmail, H ;

Celeux, G ;

Raftery, AE ;

Robert, CP .

STATISTICS AND COMPUTING, 1997, 7 (01) :1-10

[6]

Bouguila N, 2005, LECT NOTES ARTIF INT, V3587, P42

[7] Unsupervised learning of a finite mixture model based on the Dirichlet distribution and its application [J].

Bouguila, N ;

Ziou, D ;

Vaillancourt, J .

IEEE TRANSACTIONS ON IMAGE PROCESSING, 2004, 13 (11) :1533-1543

[8]

Bouguila N, 2003, LECT NOTES ARTIF INT, V2734, P172

[9]

FRIGUEIREDO MAT, 1999, LECT NOTES ARTIF INT, V1654, P54

[10]

Graybill F. A., 1983, Matrices with Applications in Statistics

← 1 2 3 →